Home Propeller Head Plaza

Technical and scientific discussion of amps, cables and other topics.

RE: A few observations (and the same conclusion)...

Dear bjh;

Your objection to our paper seems fairly typical of the response of a certain group of posters here, so I’ll answer it in detail and let this stand for the rest.

The “woeful lack of detail” you complain about refers, as far as I can tell, to our failure to cite the brands and models of the components we used. I hope you’ll pardon me if I guess that you don’t read refereed journals much, if at all. They don’t really have brand names in them, and if you put that information in, the editors take it out, since the publishers aren’t in the business of promoting individual companies in this way. (High-end audio magazines are in the business of pushing product, first, last and always, so they supply endless detail to allow their readers to imagine the alleged sublime and subtle details of the writer’s listening experience.)

So, you were clearly going to object unless we used nothing but tweako-approved gear throughout. Serious audio research, though, proceeds on an assumption that is very unpopular in these quarters but nonetheless true: Any competently designed and built player (or preamp or other line-level component) will present the recording properly, and will sound like any other such component to a degree that they cannot be told apart in a blind test, unless one of them has audible frequency response errors. Power amps are largely devoid of a recognizable signature as well, though under special conditions it can be quite easy to tell them apart. (See http://www.soundandvisionmag.com/features/651/the-ampspeaker-interface.html for proof that sometimes it’s so easy even I can do it, and what those conditions were.)

What I’m saying will seem quite radical to you, but it implies that, for example, it was OK to use the mid-priced Yamaha because neither you nor anyone else can tell the difference in a level-matched blind test between it and the $2000 Sony ES model you thought we should have used. (The level matching must be very precise and the experiment must be double blind, though no additional equipment is required.)

This will seem absurd to you, perhaps. What are all those reviewers in Stereophile and TAS writing about? I wouldn’t care to guess, but if the differences are as easy to hear as they say, it should be trivially easy to set up a proper test and prove it. In decades of going about their business, not one of them has ever done so. Instead, they have cut down whole forests to print endless tortured arguments why they should not be required to demonstrate their ability.

As part of our experiments, we located an audiophile system with an impeccable pedigree, set up our test gear in such a way as not to disturb its operation (we put a mercury relay of 0.2 ohm resistance and 18 inches of wire in series with the output of the guy’s $4000 player), and let the owner and his audiophile friends sit back and take their time with our test for an entire evening. They picked their best-sounding recordings. They turned the system up, and down. They chose what they thought were the most revealing selections, and individual passages. They could repeat a single passage and listen to it while listening to A, B and X (this is the most powerful way to hear differences, I’ve found).

We would have liked very much to discover that with certain equipment, source material and listeners that there was an audible difference. That would have made a much more interesting paper than one with (at least on music at what we established as a normal to loud playback level) null results. We searched diligently for such a combination, and didn’t find one. Many subjects were audio professionals, who make a living with their ears and who demonstrated remarkable aural prowess at times.

One reason it would have been more interesting is that null results are far less powerful than positive ones. Positive ones pretty much prove audibility, but negative ones can’t in the same way. We had some positives, as you know if you’ve read the paper. (Have you, or are you still flying blind here?) With the gain up, and the player turned off, we could hear a difference easily in the double-blind test, because the CD link was noisier. That’s been proven now, and it wasn’t hard. With the vast majority of recordings, the noise floor is above this level, so you can’t do it at any gain setting, because the background noise sounds the same either way. But we did find one recording (listed in the supplement) that is quiet enough that you can play the recording, turn up the gain, and spot the CD link. That was quick and easy to prove as well.

But what about the null results with the vast majority of high-bit recordings, played at non-stentorian levels? Maybe all our systems, including the $100,000 audiophile special, sucked. Maybe our group of recording engineers, audio degree candidates dedicated hobbyists and civilians were all hearing-impaired compared to you – every last man, woman and child. That’s pretty much what would be required to negate our results, because if even one person, with one musical selection, could do it, we’d have tested that person in detail, and written about it. If all that is true, someone else can set up a proper blind test, get positive results, and that will be that. Meanwhile, we used lots of really good equipment that was set up by people who knew what they were doing, and lots of acute listeners, and tried really hard, and came up empty. We did a good experiment and got an important result, and no amount of hand-waving or nasty name-calling of us or of the full-timers who inform and run the AES will change that. – E. Brad Meyer



This post is made possible by the generous support of people like you and our sponsors:
  Parts Connexion  


Follow Ups Full Thread
Follow Ups

FAQ

Post a Message!

Forgot Password?
Moniker (Username):
Password (Optional):
  Remember my Moniker & Password  (What's this?)    Eat Me
E-Mail (Optional):
Subject:
Message:   (Posts are subject to Content Rules)
Optional Link URL:
Optional Link Title:
Optional Image URL:
Upload Image:
E-mail Replies:  Automagically notify you when someone responds.